Performance of the pile up jet identification in CMS for Run 2 (DP-2020/020)

Link to DPS note on CDS

Introduction

The excellent operation of the LHC during the Run 2 data taking (2015-2018) has allowed the machine to reach a peak instantaneous luminosity delivered to ATLAS and CMS as high as 2x1034 Hz cm-2, leading to a large number of proton-proton inelastic scatterings to occur during the same bunch crossing ("pile up"). For the 2017 and 2018 datasets, this results in an average pile up of 32 [1]. A major impact of the pile up is to generate extra jets in the detector, in addition to those associated to the hard scattering under study. Various mitigation techniques exist to identify and reject such pile up jets [2]. The most commonly used algorithm is the "Charged Hadron Subtraction" (CHS), which discards all charged constituents associated to a vertex other than the leading vertex before starting the jet clustering procedure. The resulting jet energy is then corrected for the average energy associated to neutral particles from pile up. This procedure ensures an almost 100% efficiency for hard scatter jets. In order to further reject pile up jets, a pile up identification (PU ID) discriminator is built using a boosted decision tree (BDT), trained with 12-15 input variables characterizing the jet and the event. This note presents the performance of the discriminator that has been trained on simulated samples representing the 2016, 2017 and 2018 detector geometry and data taking conditions. It highlights the improvement brought by the Phase 1 upgrade pixel detector installed in the beginning of 2017 and that allowed to extend the tracking coverage from |𝜂|~2.5 to |𝜂|~2.7. The note finally illustrates the importance of such a PU ID even for jets inside the tracking coverage by studying the central jet multiplicity in events containing a leptonically decaying Z boson.


References
[1] The CMS Collaboration, Public CMS Luminosity Information, https://twiki.cern.ch/twiki/bin/view/CMSPublic/LumiPublicResults
[2] The CMS Collaboration, Pileup mitigation at CMS in 13 TeV data, arXiv:2003.00503, Submitted to JINST

Efficiency and mistag rate

The performance of the PU ID are assessed using three simulated samples of Drell-Yan events simulated respectively with the 2016, 2017 and 2018 detector geometry and data taking conditions. The event selection requires two opposite-charge reconstructed leptons passing identification and isolation requirements with an invariant mass between 70 and 110 GeV. The jets considered are CHS jets reconstructed with the AK4 algorithm with a transverse momentum (pT) between 20 and 50 GeV. CHS jets are considered that do not geometrically match with leptons (ΔR>0.4). They are also required to pass basic identification criteria to reject detector noise. Hard scatter jets are defined as jets geometrically matched to a generator level jet from the hard scattering with pT> 8 GeV. Pile up jets are defined as all the jets failing that condition. Four |𝜂| bins are considered separately. They correspond to the bins used during the training procedure. The efficiency (mistag rate) is defined as the fraction of hard scatter (pile up) jets for which the PU ID BDT score is greater than a given value.

Jet multiplicity in Z+jets events

The same definition as above holds, except that only jets with |𝜂|< 2.5 are considered and that no upper pT cut is applied.


Figure Caption

The PU ID mistag rate as a function of its efficiency for jets with |𝜂|<2.5 and pT∈[20,50] GeV in a Z+jet event sample, for the 2016, 2017 and 2018 simulated samples. The small variations in performance across years are found to be within the uncertainties associated to the jet energy scale and the pile up profile.
[Get pdf version][Get png version]

Figure Caption

The PU ID mistag rate as a function of its efficiency for jets with 2.5<|< 2.75 and pT∈[20,50] GeV in a Z+jet event sample, for the 2016, 2017 and 2018 simulated samples. For a given efficiency the strongly reduced mistag rate observed in 2017/2018 compared to 2016 is obtained thanks to the upgraded pixel detector that extends the tracking coverage up to |𝜂|~2.7. The variations in performance between 2017 and 2018 are found to be within the uncertainties associated to the jet energy scale and the pile up profile.
[Get pdf version][Get png version]

Figure Caption

The PU ID mistag rate as a function of its efficiency for jets with 2.75<|< 3 and pT∈[20,50] GeV in a Z+jet event sample, for the 2016, 2017 and 2018 simulated samples. For a given efficiency the strongly reduced mistag rate observed in 2017/2018 compared to 2016 is obtained thanks to the upgraded pixel detector that extends the tracking coverage up to |𝜂|~2.7. Jets in this |𝜂| region are centered outside the tracking coverage but still contains a significant number of charged hadrons. The variations in performance between 2017 and 2018 are found to be within the uncertainties associated to the jet energy scale and the pile up profile.
[Get pdf version][Get png version]

Figure Caption

The PU ID mistag rate as a function of its efficiency for jets with 3<|< 5 and pT∈[20,50] GeV in a Z+jet event sample, for the 2016, 2017 and 2018 simulated samples. The small variations in performances across years are found to be within the uncertainties associated to the jet energy scale and the pile up profile.
[Get pdf version][Get png version]

Figure Caption

The normalized central jet multiplicity in a Z+jet event sample simulated with the 2018 conditions. Jets are required to have |𝜂|<2.5 and pT> 20 GeV. The yellow histogram represents the subset of events where all reconstructed jets are geometrically matched to a generator level jet, while the red histogram corresponds to events in which one or more jets arise from pile up.
[Get pdf version][Get png version]

Figure Caption

The normalized central jet multiplicity in a Z+jet event sample simulated with the 2018 conditions. Jets are required to pass a PU ID working point corresponding to a 95% efficiency and to have |𝜂|<2.5 and pT> 20 GeV. The yellow histogram represents the subset of events where all reconstructed jets are geometrically matched to a generator level jet, while the red histogram corresponds to events in which one or more jets arise from pile up.
[Get pdf version][Get png version]

Figure Caption

The ratio of the total number of jets with |𝜂|<2.5 and pT> 20 GeV over the corresponding number of hard scatter jets before (yellow) and after (black) applying the PU ID working point corresponding to a 95% efficiency. Also shown are the ratios for PU jets only before (green) and after (red) applying the same PU ID condition.
[Get pdf version][Get png version]

Topic attachments
I Attachment History Action Size Date Who Comment
PDFpdf NewDPNote.pdf r7 r6 r5 r4 r3 manage 305.9 K 2020-04-23 - 14:34 LaurentThomas  
PDFpdf ROC_PUID_eta0to2p5.pdf r1 manage 35.1 K 2020-04-02 - 19:37 LaurentThomas  
PNGpng ROC_PUID_eta0to2p5.png r1 manage 22.6 K 2020-04-02 - 19:34 LaurentThomas  
PDFpdf ROC_PUID_eta2p5to2p75.pdf r1 manage 31.3 K 2020-04-02 - 19:37 LaurentThomas  
PNGpng ROC_PUID_eta2p5to2p75.png r1 manage 23.3 K 2020-04-02 - 19:34 LaurentThomas  
PDFpdf ROC_PUID_eta2p75to3.pdf r1 manage 27.7 K 2020-04-02 - 19:37 LaurentThomas  
PNGpng ROC_PUID_eta2p75to3.png r1 manage 22.0 K 2020-04-02 - 19:34 LaurentThomas  
PDFpdf ROC_PUID_eta3to5.pdf r1 manage 24.7 K 2020-04-02 - 19:37 LaurentThomas  
PNGpng ROC_PUID_eta3to5.png r1 manage 20.7 K 2020-04-02 - 19:34 LaurentThomas  
PDFpdf jetmultiplicity_ALL.pdf r1 manage 17.0 K 2020-04-02 - 19:37 LaurentThomas  
PNGpng jetmultiplicity_ALL.png r1 manage 19.1 K 2020-04-02 - 19:07 LaurentThomas  
PDFpdf jetmultiplicity_PassPUID.pdf r1 manage 17.1 K 2020-04-02 - 19:37 LaurentThomas  
PNGpng jetmultiplicity_PassPUID.png r1 manage 20.0 K 2020-04-02 - 19:07 LaurentThomas  
PDFpdf jetmultiplicityvsPU.pdf r2 r1 manage 24.5 K 2020-04-03 - 09:48 LaurentThomas  
PNGpng jetmultiplicityvsPU.png r2 r1 manage 25.0 K 2020-04-03 - 09:48 LaurentThomas  
Edit | Attach | Watch | Print version | History: r10 < r9 < r8 < r7 < r6 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r10 - 2020-05-29 - AlejandroGomez
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    CMSPublic All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback